Programmatic visuals: jitter/vvvv/processing? Books?

Hi all,
I’m probably diving into programmatic visuals in the upcoming months. My goal at this point is to be able to create audio or midi driven visuals from my live set. Nothing fancy, more of a “proof of concept” thing, so in the future I can create a tighter integration between my live set and the visuals, avoiding (or complementing?) working with a VJ.

My main candidates for now are jitter, vvvv and processing, as I’ve seen some pretty interesting stuff made with all of them. Jitter seems really cool, and having it work under MAX is nice because it would reinforce my investigation in that tool. Cons I can think: it’s paid. I don’t know how good (or bad) it behaves under Windows. I could go to the PD+GEM way, but PD seemed quite user unfriendly for my tastes. Not even close to MAX yet.

vvvv seems nice too, and for what i see they care about documentation and the look and feel of the app, not to mention that they are currently using it integrated with live (see OSCGlue screenshots at their site) which is my main sequencer. it’s also designed for Windows from the ground up. Don’t know how powerful it is, how reliable, etc. Oh and it’s free . Would love to hear some real world input from you guys.

And then there’s processing. Programming in Java is cool to me, as It’s a language I already know. I know of a few people working with it locally and that’s a good point for sure. Free also.

For now i would be inclined to go vvvv. Similar to jitter, but free, with a growing user community. What are your opinions regarding that? What else should I take into account?

And what books do you recommend reading for digital image processing to complement learning a particular tool?

I’d love to hear what are your experiences…

Welcome nickdatz,

while I fiercely advocate using vvvv (I couldn’t avoid it, being part of the vvvv group) I’d feel I need to correct a few of your assumptions:

OSCGlue - written by Sebastian Oschatz, also lurking here on the forum - does not need vvvv to run. You could integrate Ableton Live with Jitter using it, as well.

Free vs. Commercial License - VVVV is free for non-commercial use. This would include about 85% of all VJs or people creating club visuals, but as one can see on the licensing page, if you make a living of it, you need to pay for VVVV, too.

User base - that’s hard to guess. Since the last release, july 27th, there have been about 6000 downloads of vvvv. I’d assume that Max/Jitter have a far larger user base. Then again, that’s not necessarily an indicator of what’s better for you, or which community is more responsive.

performance/stability - we are continuously using VVVV for installations performing in front of large audiences. While VVVV is in a permanent beta state, we totally depend on it (check out our projects in the galeria blog). There may be functions which don’t work as expected, but typically those errors are easier to find and to avoid than in textual programming languages, and if a patch does run then you can usually rely on it.

As to your idea of where to use it, VVVV was practically made for exactly this. The runtime design feature (program while your patch is running) is beatifully productive for visual experimentation. Performance is typically very high, especially to non-hardware-accelerated languages like flash.

My favorite cons of VVVV?

  • doesn’t run in a browser
  • not cross platform (yet) (oh, just wait a few years. Maybe 2008)
  • doesn’t create compiled executables (not easy to make distributable binaries with it)
  • no strong audio functions

Now I’d love to hear other peoples experiences!

Thanks for the quick reply max!

As for OSCGlue, you’re totally right, what I wanted to express is that there’s people who are already using my suggested combo (ableton live + vvvv) and that’s always a good thing. Maybe a silly kind of confidence, but, you know :)

Regarding pricing, I think the licensing agreement is fair. I don’t think I’ll make a single dime on the short-to-mid-term out of my live set, so in practical terms it’s free for me to use by now. Nevertheless I would be more than happy by supporting the product when (if) the time comes.

As for stability, I’ve seen that you really use your product, and that is always a good sign.

And about no strong audio functions, what kind of integration can I expect? Would it be possible to control it via midi? Or via real time audio analysis (frequency, dynamics)?

My main working domain is audio and I’m also a systems analyst, how difficult is it for a guy like me to get interesting (whatever that means) results using vvvv? How long would it take, and how much technical info would it be advisable for me to learn first? How steep is the learning curve? How much time should I expect to “play” with the tool before going out with it?

Don’t know how powerful it is, how reliable, etc. Oh and it’s free . Would love to hear some real world input from you guys.

Well, I am not from the VVVV crew, but here in Holland we have build a club that is build (patched) and running 100% on VVVV software. And it has not crashed 1 time during an evening!! So the software is stable. (it only crashes for me when I am patching, still some buggy nodes out there)

The power… well, since VVVV is the only ‘language’ I know, and I can make nice visuals, interfaces, and just look at the impressive list off supported hardeware protocols (MIDI, DMX, RS232 etcc…) wich are all very easy to use, I would say it is prety powerfull.

Basicly I always say: if windows can do it, vvvv can do it…


Once you followed the basic tutorials, you are ready to build what you want. I remember I learned it by playing arround with the patches in the Girlpower folder. How ever, since there are so many nodes, I can tell you that you will be spending the night patching something that is already in a node. So my nr. advise is: ask in the forums!! (we are all eager to proof our skills, and helping you in the process ;) )

Thanks for sharing guys. As for technical background, any related lectures you may recommend?


i’m using vvvv for some years now and i really like it, though i couldn’t say it’s 100% stable. even when not patching i sometimes have troubles with the system but not more than with any other windows app.
what i know from max/jitter, it seems to be a tiny bit more stable, but only when working on mac. using jitter on pcs seems to make troubles.

when doing visuals with videofootage only (and not knowing how to write shaders) jitter might come more handy, since it has more imageprocessing nodes native.
in case you want to do generative visuals, vvvv is (fore sure) the better choice. vvvv’s unique feature ‘spreads’ are perfect for this workflow (although it seems like jitter 1.6beta has something alike implemented).

using controllers via midi or osc i think both are quite equal.

as for learning, i think the girlpower folder is doing well enough, in combination with constantly hitting ‘F1’ for help you’ll get into it really quick.

I would like to get deeper into the question PD/Gem vs. vvvv.

I really would like to go with vvvv, but (unfortunately) I’m on a G4 Powerbook. That means I have not enough power to VirtualPC or bootcamp Windows to run vvvv on it.

Right now I’m playing with PD/Gem and it couldn’t kill my desire for knowing vvvv better yet. But how far can I go with PD, how far could I go with vvvv.

I just would like to know some opinions from people who actually worked with both, or heared about. You know, advantages and disadvantages and such. Thank you so much!


hi, the super duper main advantage of vvvv against pd/gem for me is the comfortable GUI. as i am doing larger projects with vvvv, i would die if i had to use PD… believe me, once you used vvvv after pd, you will never switch back, if you dont have to render complex audio stuff. just to mention one thing, if you save your patch in vvvv and reopen it, all values/states will be the same. in pd every input pin will be set to its default…

besides that, if you like to do realtime graphics, vvvv has lots more features to offer…


maybe you like to try out quartzcomposer, part of the developmentevironment shipped free with newer macos.
would love to try it out but havent got a mac ;)

and if you try out tell me your experiences…

I’ve used quartz composer, it intresting, but missing too many functions I’ve got used to having! The idea is that it integrates into xcode so you build the stuff it doesnt have in c then pass it on to quartz from there, not being a coder I’m lost!
Also it has a colourfull cuddly interface that takes up too much space and isnt as easy to use or read as the greyness that is vvvv!
Running it on my macbook, I feel vvvv has more ooompf too, not masses more but certainly a few more intances on screen at once. Its bigest advantage is it is integrated into quicktime, so any app that plays quicktime can play quartz files, saving you from building intefaces for patches!

Having used PD/Gem and Max/Jitter before I have to say: vvvv is better for fast prototyping and better for large projects than every other software I have tried. And it’s more stable, once it’s stable. While developing crashes happen quite regulary, but once it runs, it runs for years.

And it’s beginner friendly. Not at first glance (the interface is really quite gray) but after having a few nodes connected, even a newbie can get nice looking results much, much faster than with other programming languages.

As tonfilm said: If you don’t need much audio, vvvv is the weapon of choice.

Thanks alot, mates. That’s good info so far. So I’m going to stick with Quartz Composer and some PD/Gem until I get my hand on some decent windows machine, or if money alows some intel mac (bootcamp native windows).

And yes, if my opinion counts for you elektromeier, I’m probably going to send you some messages, if that is ok for you, about my working exprience with QC.